Drug addiction can be conceptualized as an extreme form of memory, a memory that is distinctly long-lasting and completely new for individuals who have never experienced drugs of abuse. Whereas modification of pre- existing synapses/neural circuits has been widely accepted as a cellular mechanism for forming memories, we hypothesize that drug addiction as a new and extremely robust form of memory is mediated not only by modification of the existing synapses, but also by the formation of new synaptic connections and thus new neural circuits. This hypothesis, although seemingly striking, is consistent with several lines of evidence showing that following exposure to cocaine or other addictive psychostimulants, new dendritic spines and premature excitatory synaptic connections are formed in the nucleus accumbens (NAc), a critical brain region for the development of drug addiction. To directly measure the potential new synaptic connections/neural circuits induced by exposure to cocaine, we propose to adapt and develop an innovative imaging technique, the GRASP (GFP reconstitution across synaptic partners) technique, which can label newly formed synaptic connections in vivo. Our strategy is to split GFP into two halves, one expressed in potential presynaptic terminals and the other in potential postsynaptic terminals. These two halves do not fluorescence individually but will reconstitute into a fluorescent GFP when the pre- and post-synaptic terminals interact to form new synapses. The proof-of-concept data from our preliminary studies suggest that this experimental strategy is highly promising. In this application, our first objective is to optimize the GRASP technique such that it can be readily used for in vivo labeling of new excitatory synapses. Our second objective is to use this technique to characterize a potential new neural projection to the NAc following cocaine exposure. Using a Fluoragold- based tracing approach, our preliminary data showed that following exposure to cocaine, the NAc received new, intense innervations from the lateral habenula, a projection that was not observed in saline-treated control animals. Thus, the proposed work will not only develop an in vivo GRASP technique, but also use this technique to address an important neuroscience question. As such, this proposal is consistent with the mission of CEBRA application, and the outcome of this proposal will provide the field with a novel in vivo technique and a potentially novel circuitry-based mechanism underlying cocaine addiction.